Comparison of pre- and post-processors for ensemble streamflow prediction

Abstract

This study conducted a broad review of the pre- and post-processor methods for ensemble streamflow prediction using a Korean case study. Categorical forecasts offered by the Korea Meteorogical Administration and deterministic forecasts of a regional climate model called Seoul National University Regional Climate Model(SNURCM) were selected as climate forecast information for the pre-processors and incorporated into Ensemble Streamflow Prediction(ESP) runs with the TANK hydrologic model. The post-processors were then used to minimize a possible error propagated through the streamflow generation. The application results show that use of the post-processor more effectively reduced the uncertainty of the no-processor ESP than use of the pre-processor, especially in dry season.